Irrigation
M. Komasi; A. Alizadefard
Abstract
Introduction: The occurrence of successive droughts, along with increasing water needs and lack of proper management of water resources has caused a water crisis that has various environmental and economic consequences. In addition to the drought, the change in the cropping pattern towards water crops ...
Read More
Introduction: The occurrence of successive droughts, along with increasing water needs and lack of proper management of water resources has caused a water crisis that has various environmental and economic consequences. In addition to the drought, the change in the cropping pattern towards water crops has also made the water crisis the first critical phenomenon in recent years in the community, which has a direct impact on the agricultural sector as the largest consumer of water. Therefore, optimizing the cropping pattern is one of the most important factors in managing water resources and coping with water shortages. In this study, to determine the optimal cropping pattern of major crops in Silakhor plain in the next three years using two approaches using Linear Programming and Meta-Heuristic Algorithms. Materials and Methods: In the first step, in order to determine the optimal cropping pattern with the aim of maximizing farmers' incomes in the next three years and the limited water and land available, the amount of rainfall recharge is used as a criterion to determine the water exploitation interval and determine the minimum and maximum exploitation each year. In order to forecast rainfall, SARIMA time series models and Genetic Programming were used considering the data of the last 10 years in both seasonal and monthly modes, and according to RMSE and D.C. criteria, a better model was selected. Then, for each crop year, 100 exploitation scenarios were determined according to the amount of groundwater recharge caused by rainfall and the amount of exploitation in previous years. In the second step, Linear Programming was used to determine the optimal cropping pattern with the aim of maximizing farmers' incomes and limitations of exploitable water in each scenario and arable land. The price of each product is projected according to the average long-term inflation of the country, i.e., 20%, and the profit from the cultivation of each product was calculated as a proportion of the price of the product in each year by examining the previous years. Finally, the performance of three types of Static, Dynamic, and Classified Dynamics Penalty Functions into two algorithms, Differential Evolution and PSO was investigated to achieve the results obtained from Linear Programming. Static penalty functions use a constant value during the optimization process, whereas in dynamic penalty functions, the fines are modified during the process and depend on the number of generations. In the classified dynamics penalty, groups of violations are also determined, and the penalty of each response is determined according to the amount of violation of the restrictions and the generation number. Results and Discussion: The results show that with increasing groundwater exploitation, farmers' incomes also increase; However, in the exploitation of more than 223.5, 222.2, and 225.1 million cubic meters for the cropping years 2020-2021, 2021-2022, and 2022-2023, respectively, the limitation of the total arable land has prevented the increase of the area under cultivation, and by increasing exploitation, farmers' incomes remain stable. Also, in order to cultivate four crops of wheat, barley, rice, and corn with the current area under cultivation in Silakhor plain, 142 million cubic meters of water is harvested annually from underground sources. By optimizing the cropping pattern for the four crops studied, with the current water exploitation, the income of farmers in the region will increase by 18%. In general, the PSO algorithm answers this problem much faster. The average number of iterations of the PSO algorithm to solve each scenario in this problem is 38% of the number of iterations of the Differential Evolution algorithm. Overall, in solving this problem, the PSO algorithm has performed better in 84% of the scenarios. In penalty functions, the best performance in both algorithms belongs to the classified dynamics, dynamic, and static penalty functions, respectively. By changing the penalty function from static to classified dynamics penalty function, the number of iterations of the Differential Evolution algorithm to achieve the Linear Programming solution is reduced by an average of 11%; In contrast, the PSO algorithm did not react significantly to the change in the penalty function, and its repetitions decreased by an average of only 3%. Conclusion: The results show that the cropping pattern of the region is not optimal, and with the increase of water exploitation, it will move towards the cultivation of water products. Also, by optimizing the cultivation pattern of the region, farmers' incomes can be increased. Examination of Differential Evolution and PSO algorithms with three types of penalty functions also show that using the classified dynamics penalty function in the PSO algorithm can have good results.
